At first we where using a 24V power supply with sense wires that should bring the voltage up to compensate for the drop of voltage caused by the cable length. This didn’t work very well as the power supply wasn’t able to handle to voltage drop propoperly (when using the ROV at full throttle). Our new plan is to use a 60V DC power supply and a DC-DC converter with 12-60V input and a adjustable output (around 18V). As things have been a bit busy the last few weeks we havent been able to work on this yet.
Let me know if it works, I hope the DC-DC converter will remain stable under those loads… I’m just looking around to see if anyone got it working yet, should be possible but maybe the investment in a stronger supply is indeed the way to go.
Hi Klaas, we currently have a working setup to at least power the electronics seperately over POE through the 75m phantom tether. We removed the Phantom-X communication boards on both ends, and just put a POE splitter in the ROV. This allows us to power the Raspberry and pixhawk with 5v (also replaced the 5v on the pixhawk rail for the camera servo with this, as we wanted to separate thruster power supply from electronics). Also a note: we can still read the V and A of the battery, but we switched the power for the pixhawk on that connector as well. Communication is now always stable, even when the motor voltage drops. Our next step is to completely replace the battery by a cable as well, to power just the motors and the lights. We will try a separate cable first, maybe later upgrade to ethernet+power in one cable, though I suspect we may see some signal degradation…I will keep you updated!
We’ve got the external power supply working (a 80V 40A one).
Seperate 4 wire cable for the power supply. We use it without the sense, as it can only correct about 2,5V if the voltage drops. We put about 20V on the input over a 50m cable. At gain=100%, max thrust down and forward, it seems to reach our max of 40A, and the voltage still drops some, but QGroundcontrol reports 7V remains at the ROV end, enough to keep the ESCs working fine. We do see the lumens losing some brightness at this point, and I suspect some thruster power is lost, but since these are extreme test scenarios, we should be fine for normal operation. Even if the voltage would drop lower, we still have an active connection due to the separate POE of the raspberry/pixhawk I mentioned before. We are now working on adding some floaters to the heavy power cable so we don’t impede the ROV with that weight. Good luck with your setup! I hope @bluerobotics may also be interested in producing a BlueROV version that runs power through cables as well, as this increases the maximum dive time for users.
Bart Arts, if you’re getting voltage drop down to 7V, then you’ll be losing most of the thruster power. At 7V you probably get about 1/4 of the thrust at 16V. The lumens will also start to become dimmer under 10V. Regardless, it’s very cool that you’ve got this working! Very nice.